Metoda grupiranja kupaca električne energije u karakteristične skupine potrošnje

doctoral thesis

• Title (english): Method for Electric Power Consumers Clustering into Representative Consumption Classes
• Author: Dino Mileta; MBZ: -823
• Mentor: Zdenko Šimić (mentor)
• Committee member: Zdenko Šimić (predsjednik povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; (Department of Energy and Power Systems); Zagreb
• Defense date and country: 2013-02-28, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering
• Universal decimal classification (UDC): 621.3 - Electrical engineering

Abstract

Za uspješno održavanje ravnoteže proizvodnje i potrošnje u elektroenergetskom sustavu potrebno je poznavati značajke potrošnje kupaca te dobro predviđati njihovu potrošnju. Poznavanje krivulja opterećenja kupaca u vremenu obračunskog razdoblja moguće je korištenjem naprednih brojila, s mogućnošću mjerenja i pohrane krivulje opterećenja, na obračunskim mjestima kupaca. Opremanje obračunskih mjernih mjesta svih kupaca takvim brojilima još je uvijek vrlo skup, složen i za određene kategorije kupaca ekonomski neopravdan postupak. Stoga se za neke od tih kategorija kupaca moraju koristiti drugačije metode pridjeljivanja nadomjesnih krivulja opterećenja. Osim opskrbljivača korisnici nadomjesnih krivulja opterećenja u tržišnim uvjetima i liberaliziranom tržištu mogu biti operatori i regulator koji mogu koristiti nadomjesne krivulje opterećenja kod modeliranja vrste potrošača na tržištu kod izrade tarfinog sustava za prijenos ili distribuciju. Osim na tržištu električne energije krivulje opterećenja mogu se koristiti i kod planiranja mreža. Ako te mreže napajaju kupce bez intervalnih brojila, opet se javlja potreba za modeliranjem pomoću nadomjesnih krivulja opterećenja. Takvi dijagrami, dobiveni pomoću nadomjesnih krivulja opterećenja, mogu elektroenergetskim planerima olakšati procjenu gdje treba graditi nove distribucijske kapacitete te pomoći u ocjeni mogućnosti priključivanja distribuirane proizvodnje na distribucijsku mrežu. U ovom doktorskom radu predstavljen je razvijeni model za prepoznavanje karakterističnih skupina kupaca električne energije, određivanje njihovih značajki i broja skupina s primjenom na stvarni elektroenergetski sustav. Za grupiranje je korišten algoritam za hijerarhijsko grupiranje i k-means algoritam. Osim prepoznavanja postojećih karakterističnih skupina kupaca električne energije predstavljen je i model za klasifikaciju novih kupaca koji nemaju mjerenje krivulje opterećenja u karakteristične skupine kupaca prepoznate grupiranjem. Kao metoda za izradu klasifikacijskog modela korišteno je stablo odlučivanja. Za verifikaciju čitavog modela provedena je simulacija na virtualnoj transformatorskoj stanici koja je uspoređena sa stvarnim mjerenjima.

Abstract (english)

The market operator and the transmission system operator play a crucial role in the electric power system operation, in regard to achieving a balance between consumption on the one hand, and production within the system and the inflow of electric energy from outside of the system on the other hand. The market operator mainly does this by coming up with a market plan based on the contractual schedule of participants on the electric energy market, while the transmission system operator does this by providing services and by coming up with the system’s operation plan. The process of liberalization and the emergence of new technologies, placing great emphasis on understanding and identifying customer in the electricity sector. In the electricity sector, a load profile is a curve showing the amount of electric energy customers’ use over a period of time. This is important because in retail energy markets supplier’s obligations are settled on an hourly or sub-hourly time basis. For most customers consumption is traditionally measured on a twice yearly or even yearly basis by meter readings. Standard load profiles are used to convert this monthly or sometimes yearly consumption data into estimates of hourly or sub-hourly consumption in order to determine the supplier’s obligation. For each hour, these estimates are aggregated for all customers of an energy supplier, and the aggregate amount is used in market settlement calculations as the total demand that must be covered by the supplier. Data regarding load profiles of all customers during balancing settlement period is attainable by placing meters that have the ability to measure and store load profiles, in accounting metering points (OMM). Supplying the accounting metering points of all customers with such meters is expensive, complex and for certain customer groups a financially unjustifiable move. Hence, for certain customer groups in the Republic of Croatia, the supply of accounting metering points with equipment that allows the measurement and storage of load profiles is not possible, thus, a characteristic load profile (CLP) is used instead. Theme of the doctoral dissertation is the creation of framework to identify typical groups of customers of electricity, determining their characteristics and the number of groups with application to the Croatian electricity power system. Method is developed for predicting energy consumption, creation of characteristic customer groups load profiles and classification of new customers. These efforts should contribute to the quality of power system planning and reducing market risk of the supply of electricity to end customers. In the first phase, data is collected from household, entrepreneurs and public lighting customers. The second phase, data pre-processing, contains four subphases: analysis, verification, transformation and integration. The third phase, data mining, concentrates on the framework that clusters and classifies customers according to their consumption curve. Classification is made by applying a decision tree. The classification module focuses more on using the knowledge gathered from questionnaires for classification of customer classes. Finally, the last phase verification of the entire model is implemented on the virtual substation and then compared with the actual measurements.

• Keywords: tržište električne energije; nadomjesne krivulje opterećenja; grupiranja kupaca električne energije; klasifikacija kupaca električne energije; modeliranje opterećenja; energija uravnoteženja; mjerenje potrošnje; karakteristične skupine kupaca
• Keywords (english): electricity market; clustering; load profiles; electricity market; electricity settlement; load modeling; classification; characteristic load profile
• Language: croatian
• URN:NBN: urn:nbn:hr:168:463945
• Project: Number: 036-0361590-1591; Title: Razvoj alata za analizu tržišta električne energije; Leader: Slavko Krajcar; Jurisdiction: Croatia; Funder: MZOS; Funding stream: ZP
• Study programme: Title: Doctoral study programme "Electrical Engineering and Computing"; Study programme type: university; Study level: postgraduate; Academic / professional title: doktor/doktorica znanosti, po-dručje tehničkih znanosti (doktor/doktorica znanosti, po-dručje tehničkih znanosti)
• Catalog URL: http://lib.fer.hr/cgi-bin/koha/opac-detail.pl?biblionumber=41196
• Type of resource: Text
• Extent: 239 str ; 30 cm
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2019-11-05 12:43:40